Buy now, pay later is credit now: what the new rules mean

Photo: rawpixel via rawpixel (CC0 1.0)
Four payments of $47.25, no interest, no credit check, and for the best part of a decade no regulator either. Buy now, pay later grew up in a gap in the National Consumer Credit Protection Act that exempted short-term, interest-free arrangements, and the industry built a category, two ASX darlings and one very large overseas acquisition inside that gap before the government closed it. The legislation that brought BNPL under the credit framework passed in late 2024 and the obligations commenced in June 2025. A year on, the changes are visible at checkout if you know where to look, and I’ve spent the last few months asking founders, lenders and retailers what actually moved. I spent two years inside a payments start-up that sat next to this industry, close enough to see the numbers and far enough not to be responsible for them, which is the most comfortable seat in fintech.
What changed, in one paragraph #
The Treasury Laws Amendment (Responsible Buy Now Pay Later and Other Measures) Act 2024 created a new category of regulated credit called a low cost credit contract. Providers of those contracts must hold an Australian credit licence from ASIC, become members of the Australian Financial Complaints Authority, comply with the responsible lending obligations in a modified form, meet requirements on fees and caps, and follow the same hardship, disclosure and dispute rules as other lenders. The “modified” part is the negotiation that took years: providers can choose a lighter assessment for lower limits, scaled to the amount of credit, rather than the full assessment a mortgage requires. They can no longer do no assessment at all. The ASIC guidance that followed sets out how the regulator expects the scaled assessment to work.

What it looked like from inside a provider #
The providers I’ve talked to describe 2025 as the year the product team and the compliance team merged. The affordability check had to be built, and built into a checkout flow that’s meant to take eleven seconds. The clever ones used the Consumer Data Right for exactly this (I’ve written about why that worked), pulling verified income with consent for customers above a threshold and using bureau data below it. The less clever ones bolted a questionnaire onto the front and watched conversion drop.
Hardship was the other build. A BNPL provider that had, historically, dealt with missed payments by pausing the account and charging a late fee now has to have a hardship process that would satisfy AFCA, which means people, policies and a willingness to vary the contract. Several providers told me this was the larger cost, because it’s a permanent operating expense rather than a one-off engineering job.
And the licence itself, with its responsible manager requirements, its compliance framework and its ongoing reporting, ended some of the smaller providers. A few merged. A few exited. The market that emerged is more concentrated and, honestly, more like the lending industry it always was.
What a retailer now owes its customers #
This is the part I think is underappreciated. If you’re a retailer offering BNPL at checkout, you’re not the credit provider, and the licence obligations sit with the provider. But you’re now the point of sale for a regulated credit product, and that has consequences.
The advertising rules apply to you. “Interest free” claims, the prominence of the instalment price relative to the full price, the disclosure of fees: the consumer law and ASIC’s advertising guidance apply to the retailer’s presentation of the product, and the provider’s contract with you almost certainly now requires you to follow their approved wording. The surcharging question, which the Reserve Bank has been looking at for years, is also live; whether and how you can pass on the merchant fee, which for BNPL runs well above card rates, is a rule that’s been under review and one you should have a current answer for.
And the customer complaint about a BNPL product will land, in the first instance, with you. The provider handles the credit side. You handle the “I was offered this at checkout and didn’t understand it” conversation, and the retailers that have trained their customer service teams on what BNPL now is have fewer of those conversations than the ones who treat it as a payment button.
The other adjustment providers made was to the credit limit itself. Under the old model a new customer might be offered a limit of $1,000 or more on sign-up, with the algorithm learning from repayment behaviour after the fact. Under the scaled assessment the initial limit is tied to what the provider can verify, so a customer who declines to share transaction data starts small. Several founders told me this single change did more to reduce their bad-debt line than anything in the previous five years, which raises the obvious question of why it took a law. The answer they gave, with varying degrees of candour, is that growth was being measured in customers and limits, not in loans repaid.
The fee cap that changed the economics #
The law’s fee provisions, which limit what a provider can charge if it takes the lighter assessment path, changed the unit economics of the smallest transactions. A $30 purchase split four ways was always a marginal product for the provider; with capped fees and a mandatory assessment, some providers have raised minimum transaction values or quietly steered small baskets to a card payment instead. The average BNPL basket has gone up. The customer who used the product for a $25 pair of socks has, in many cases, stopped seeing it offered.
Whether that’s a loss depends on your view of whether four payments on a pair of socks was ever a good idea. Mine, which I’ll defend, is that it wasn’t, and that the industry’s growth in the low-value segment was borrowing demand from its own future customers. The financial counselling sector, which had been saying this since 2018 and which lobbied hard for the law, has the data. Consumer advocates at organisations such as the Consumer Action Law Centre documented the pattern of stacked BNPL accounts and missed essential payments for years before the government acted.
What the customer actually sees #
If you’ve used one of these products since mid-2025 you’ll have noticed the differences even if you didn’t know why. A first-time customer now goes through a short identity and affordability step before the first purchase rather than after the first missed payment. Limits start lower and grow with repayment history, which is how credit cards have worked for forty years. The late fee is capped and disclosed up front. And if things go wrong there’s a hardship option in the app rather than a support email that goes unanswered, because the provider is now answerable to AFCA if it doesn’t respond.
The complaints data will tell the story over the next couple of years. AFCA publishes its figures by product category, and the number I’ll be watching is not the total but the proportion resolved at the first stage, because that’s the measure of whether providers built a real hardship function or a form.
The retailer’s checklist, briefly #
For a retailer that wants to keep offering instalments without inheriting someone else’s compliance problem, the work is smaller than it sounds. Get the provider’s current approved wording for every place the product appears and use it exactly. Show the full price at least as prominently as the instalment. Know your position on surcharging and make sure the checkout reflects it. Train the customer service team on the one-paragraph version of what BNPL now is. And read the merchant agreement’s indemnity clause, because several providers updated theirs in 2025 and the changes were not in the retailer’s favour.
Retailers with their own apps have one further decision: whether to keep the third-party BNPL button at all, now that the bank-issued instalment features on ordinary cards do much the same job with none of the merchant fee. Several of the larger chains I’ve spoken to are quietly running that comparison, and the answer depends on whether the BNPL provider’s marketplace is still sending them customers or just processing the ones they already had.
Where the industry goes next #
Three directions, from what I’m hearing. The larger providers are moving up the value chain into what is, in effect, a credit card by another name: longer terms, higher limits, sometimes interest. The bank-owned products, which several of the big four launched to compete, are quietly being folded back into ordinary card and personal loan offerings now that the regulatory arbitrage is gone. And a smaller group is going business-to-business, offering instalments to small businesses buying inventory, where the responsible lending framework is different and the margins are better.
The founder story here is worth telling. The BNPL boom produced some of the fastest-growing companies in Australian history and a great many pitch decks that copied them, and the lesson the venture market took from the regulation isn’t “regulation kills fintech”. It’s that a business model built on an exemption has a shelf life measured by the next review, and the companies that treated themselves as lenders from the start, with the compliance cost that implies, were the ones that didn’t have to rebuild in 2025.
Two of the founders who told me, with great confidence, at the same fintech conference in 2021 that regulation would never touch them no longer work in fintech. I don’t say that with any satisfaction. I was on a panel at the same conference saying something nearly as wrong about open banking, and the recording is, unfortunately, still online.
— Sophie Lindqvist, AI, Start-ups & Fintech Editor, Technology Spectator